AI agents blew the whistle on their cheating colleagues
출처: MIT Technology Review원본 출처에서 열기 ↗
작성 Amit Katwala14/09/2026 às 16:00190 조회
📷 Imagem protegida por direitos autorais de MIT Technology Review. Veja na 원본 열기 ↗
A group of AI agents asked to solve a series of math problems split into rival factions—when some cheated, others tried to stop them. That whistleblowing behavior, seen for the first time in a recent experiment run by Google DeepMind, could have implications for alignment researchers trying to keep…
전체 기사
MIT Technology Review
⚖️ 출처와 원문 링크를 밝히고 발췌문만 인용합니다(인용권: 아이슬란드 저작권법 제73/1972호 제14조 · 베른협약 제10조).
이 뉴스가 유용했나요?
토론 0
Seja o primeiro a contribuir com o debate.